Environmentally friendly GIS multi-dimensional optimization simulation method and system with multi-physics coupling

Abstract

A multi-physics coupled environmentally friendly GIS multivariate optimization simulation method and system is disclosed. The method includes the following steps: establishing a three-dimensional model for finite element simulation and setting material parameters and boundary conditions; performing electro-thermal-fluid multi-physics finite element simulation to obtain simulation temperature data; reducing the order of the simulation temperature data and training a reduced-order model; using the reduced-order model to generate small-error sample data for sensitivity analysis, calculating the response surface modeling priority index of each input variable, and selecting key variables; calculating the busbar material based on the selected key variables; obtaining multiple sets of simulation result data by randomly changing key parameters multiple times, fitting the response surface expression, constructing and solving the optimization objective function to obtain the optimal key variables. This invention achieves efficient synergistic optimization of environmentally friendly GIS structural parameters and thermal performance by combining multi-physics coupled simulation, reduced-order sensitivity analysis, and response surface optimization.

Description

Technical Field

[0001] This invention belongs to the field of equipment parameter simulation technology, and relates to an environmentally friendly GIS multi-element optimization simulation method and system with multi-physics coupling. Background Technology

[0002] Gas-insulated metal-included switchgear (GIS) is a complete set of switchgear and control equipment composed of electrical components such as circuit breakers, disconnecting switches, grounding switches, current transformers, voltage transformers, surge arresters, busbars, bushings, or cable terminals, arranged according to a specific wiring method, using SF6 gas or other gases as the insulating medium. SF6 gas is widely used in the power industry due to its excellent electrical insulation and arc-extinguishing properties. However, with increasing concern about the greenhouse effect and life-cycle emissions of SF6, the industry is forced to maintain the existing insulation levels and thermal reliability of ultra-high voltage equipment while reducing or even eliminating the use of SF6. Alternative media such as clean air and C4F7N differ significantly from SF6 in dielectric strength, relative permittivity, density and viscosity, specific heat capacity, and thermal conductivity. Existing empirical dimensional and local structural treatments are no longer reliable, and design work has shifted from "following the rules" to evidence-driven decision-making based on multiphysics simulations.

[0003] This transformation is particularly pronounced at voltage levels such as 550kV. The heat generated by electromagnetic losses alters the gas properties, reshaping the internal flow field and heat transfer, which in turn affects the temperature and electric field distributions. Millimeter-scale rounded radii and corrugated structures coexist with meter-scale overall machine dimensions, creating a coupling relationship across time and space scales. To capture boundary layer heat transfer, the model must employ high-quality anisotropic meshes at sharp corners, gaps, and insulating interfaces, and bear the pressure of interpolation errors and consistency management during cross-solver data interaction. The number of degrees of freedom in the discretized overall model rapidly expands, making single-case solution time and memory usage bottlenecks.

[0004] In the thermal stability optimization design of environmentally friendly GIS, design variables have expanded from gas ratio and working pressure to include the profiles of equalizing rings and electrodes, critical gaps, heat dissipation components, and insulation morphology. Objectives and constraints are intertwined between temperature rise, quality, cost, and manufacturability. Experience-based local modifications and single-parameter scans often fail to find the global optimum; extrapolation of low-order surrogate models trained with a small number of samples carries high risks, while directly embedding full-order models into global optimization incurs unacceptable computational costs.

[0005] Currently, existing GIS simulation methods include: Patent CN118446110A provides a rapid simulation method and twin system for steady-state temperature rise of GIS disconnect switches. This method achieves rapid prediction and inversion of the internal hot spot temperature of GIS disconnect switches by constructing a multi-physics bidirectional coupling model and a deep neural network model; while Chinese patent CN119720469A provides a GIS equipment simulation method and system that considers multi-physics coupling. This method uses a local simulation geometric model obtained after the GIS equipment has been deformed to perform simulation calculations on the deformed GIS equipment again, thereby realizing the bidirectional coupling of multi-physics fields between electromagnetic field, thermal flow field and stress field.

[0006] In summary, while existing GIS simulation methods involve rapid calculation of GIS temperature rise and simulation calculation of geometric deformation, they do not include rapid temperature rise calculation with geometric parameters as input and their combination with other methods to guide the optimization design of flow structure in environmentally friendly GIS. Summary of the Invention

[0007] To address the technical problems of high simulation time costs and reliance on experience in structural design, which makes it difficult to find the global optimal solution, existing technologies provide a rapid iterative multivariate optimization simulation method and system for environmentally friendly GIS using multiphysics coupling. The method includes the following steps: establishing a three-dimensional model for finite element simulation and setting material parameters and boundary conditions; performing electro-thermal-fluid multiphysics finite element simulation to obtain simulation temperature data; reducing the order of the simulation temperature data and training a reduced-order model; using the reduced-order model to generate small-error sample data for sensitivity analysis, calculating the priority index of response surface modeling for each input variable, and selecting key variables; calculating the busbar material based on the selected key variables; repeatedly and randomly changing key parameters to obtain multiple sets of simulation result data, fitting the response surface expression, constructing and solving the optimization objective function to obtain the optimal key variables. This invention achieves efficient synergistic optimization of structural parameters and thermal performance of environmentally friendly GIS by combining multiphysics coupling simulation, reduced-order sensitivity analysis, and response surface optimization.

[0008] The present invention adopts the following technical solution: The first aspect of this invention provides an environmentally friendly GIS multivariate optimization simulation method with multiphysics coupling, comprising: S1. Establish a three-dimensional model for environmentally friendly GIS finite element simulation, and parameterize the component dimensions and current magnitude; set material parameters and boundary conditions for the three-dimensional model according to the working conditions; S2. Based on the three-dimensional model and the set material parameters and boundary conditions, perform multi-physics finite element simulation of electro-thermal-fluid fields; set multiple combinations of different current magnitudes and size parameters to obtain simulation temperature data under multiple working conditions; S3. Using simulated temperature data as a sample data matrix, the sample data matrix is ​​reduced in order to generate a reduced-order data matrix, and the reduced-order data matrix is ​​used to train and generate a reduced-order model. S4. Use the reduced-order model to generate small-error sample data for sensitivity analysis, calculate the first-order sensitivity index and total effect index of each input variable, and construct the response surface modeling priority index of each input variable to screen the input variables and obtain the key variables. S5. Based on the selected key variables, calculate the busbar material; repeatedly change the values ​​of key parameters to obtain multiple sets of simulation result data, and fit the response surface expression; based on the response surface expression, determine the busbar material, average temperature rise of GIS and maximum temperature rise of GIS under different key variables, construct and solve the optimization objective function to obtain the optimal key variables.

[0009] Preferably, S3 includes: The sample point data matrix is ​​decomposed into a basis vector matrix and an original singular value matrix; Based on the original singular value matrix, calculate the energy accumulation ratio under different cutoff orders; determine the minimum cutoff order when the energy accumulation ratio is higher than the energy threshold, and use it as the final cutoff order; retain the singular values ​​of the original singular value matrix before the final cutoff order, and use them as the lower-order singular value matrix. Extract the first few basis vector matrices of the final truncation order from the basis vector matrix to form the low-order basis vector matrix; combine the low-order singular value matrix with the low-order basis vector matrix to generate the reduced-order data matrix of the low-order mode. A reduced-order model is generated by training using a reduced-order data matrix.

[0010] Preferably, S4 includes: The values ​​of the input variables in the reduced-order model are changed multiple times within a predetermined range to generate corresponding small-error sample data. Based on the small-error sample data, parameter sensitivity analysis is performed to calculate the variance contribution of each input variable to the output index, obtaining the first-order sensitivity index and the total effect index. The formula for the first-order sensitivity index is: ; In the formula, the first-order sensitivity index express The percentage of independent contributions to the overall output variance; Let i be the i-th input variable; This is the output of the reduced-order model; Indicates in fixed Under the given conditions, the conditional expectation obtained by averaging the remaining variables; symbol Represents the expectation operation, symbol Indicates variance operation; Based on the first-order sensitivity index and the total effect index, considering the magnitude of the influence of each variable on the output and the magnitude of the interaction with other variables, a response surface modeling priority index is constructed, and key variables are selected.

[0011] Preferably, the process of selecting key variables in S4 is as follows: For any input variable, the covariance between the corresponding input variable and the output of the reduced-order model is used as the numerator, and the product of the standard deviations of the corresponding input variable and the output of the reduced-order model is used as the denominator to calculate the linear correlation index between the corresponding input variable and the output of the reduced-order model; the square of the linear correlation index between the corresponding input variable and the output of the reduced-order model is used as the corresponding linear correlation determination index. Calculate the difference between the first-order sensitivity index and the linear correlation determination index of the corresponding input variable, and take the maximum value between this difference and 0 as the corresponding nonlinear effect index; calculate the difference between the total effect index and the first-order sensitivity index of the corresponding input variable, and take the maximum value between this difference and 0 as the corresponding interaction effect index. Based on the total effect index, nonlinear effect index, and interaction effect index, a response surface modeling priority index is constructed for each input variable. The input variables are then sorted from largest to smallest according to the response surface modeling priority index, and a predetermined number of input variables with the largest response surface modeling priority index are selected as key variables.

[0012] Preferably, the specific formula for calculating the response surface modeling priority index is as follows:

[0013] In the formula, c i For variables Response surface modeling priority index; Let be the interaction effect index of the i-th input variable; N i Let be the nonlinear effect index of the i-th input variable; For variables The total effect index; Representing variables The first-order sensitivity index; e This is the stabilizing factor in the denominator.

[0014] Preferably, the specific representation of the busbar material usage in S5 is as follows: ; In the formula, Materials used for busbars; The length of the cylinder; D b1 The inner diameter of the busbar; D b2This refers to the outer diameter of the busbar.

[0015] Preferably, the response surface expression in S5 is:

[0016] In the formula, This indicates the m-th output; This represents the number of key variables, which is set to 3 here. The constant term fitting coefficients are the response surface constants corresponding to the m-th output. The coefficients for the first-order term of the response surface corresponding to the m-th output are: The coefficients for the quadratic term of the response surface corresponding to the m-th output are: The coefficients for the response surface interaction term corresponding to the m-th output are: Let i be the i-th key variable corresponding to the m-th output.

[0017] Preferably, the process of constructing and solving the objective function in S5 is as follows: Set upper and lower limits for the input of key variables; take minimizing the busbar material consumption and the average temperature rise of GIS as the optimization objectives, and take the maximum temperature rise of GIS not exceeding a predetermined temperature threshold as the constraint condition; transform the multi-objective into a single-objective optimization objective function by weighting the two optimization objectives; solve the optimization objective function under the constraint condition to obtain the optimal key variables within the upper and lower limits of the input of key variables.

[0018] A second aspect of the present invention provides an environmentally friendly GIS multi-element optimization simulation system using multi-physics coupling, comprising: The 3D model building module establishes a 3D model for environmentally friendly GIS finite element simulation and parameterizes the dimensions of components and current magnitudes; based on the working conditions, material parameters and boundary conditions are set for the 3D model. The model simulation and data sampling module performs multiphysics finite element simulations of electro-thermal-fluid fields based on the 3D model and the set material parameters and boundary conditions; it sets multiple combinations of different current magnitudes and size parameters to obtain simulation temperature data under multiple operating conditions. The reduced-order model construction module uses simulated temperature data as a sample data matrix, performs reduced-order processing on the sample data matrix to generate a reduced-order data matrix, and uses the reduced-order data matrix to train and generate a reduced-order model. The key variable screening module uses a reduced-order model to generate small-error sample data for sensitivity analysis, calculates the first-order sensitivity index and total effect index of each input variable, and constructs a response surface modeling priority index for each input variable to screen the input variables and obtain key variables. The key variable solving module calculates the busbar material requirements based on the selected key variables; it obtains multiple sets of simulation result data by randomly changing the values ​​of key parameters multiple times, and fits the response surface expression; based on the response surface expression, it determines the busbar material requirements, average temperature rise of GIS, and maximum temperature rise of GIS under different key variables, and constructs and solves the optimization objective function based on this to obtain the optimal key variables.

[0019] A third aspect of the present invention provides a terminal, including a processor and a storage medium; The storage medium is used to store instructions; The processor is used to operate according to the instructions to execute the steps of the environmentally friendly GIS multi-element optimization simulation method with multi-physics coupling.

[0020] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of an environmentally friendly GIS multi-element optimization simulation method with multi-physics coupling.

[0021] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention constructs a multi-physics coupled finite element model of environmentally friendly GIS (Geometric System) based on electro-thermal-fluid fields, and parameterizes parameters such as current magnitude and structural dimensions to achieve accurate simulation analysis of GIS operation under different working conditions. Compared with traditional single-physics field analysis methods, this invention can comprehensively consider the coupling effects between electromagnetic losses, temperature rise distribution, and gas flow, more realistically reflecting the thermal characteristics and operating performance of environmentally friendly GIS under complex operating environments. This improves the accuracy and engineering applicability of simulation results, providing a reliable basis for the structural design and thermal management optimization of environmentally friendly GIS.

[0022] 2. This invention effectively reduces the modeling complexity and computational cost in high-dimensional parameter spaces by reducing the order of multi-condition simulation data and combining it with sensitivity analysis to screen key variables. By introducing first-order sensitivity indices, total effect indices, nonlinear effect indices, and interaction effect indices, a response surface modeling priority index is constructed. This index not only identifies key parameters that significantly affect the output results but also simultaneously characterizes the nonlinear effects and interactive coupling characteristics of variables. This improves the accuracy of key variable selection and the fitting efficiency of the response surface model, while reducing the interference of irrelevant variables on the optimization results.

[0023] 3. This invention establishes a response surface expression based on key variables and aims to minimize busbar material consumption and the average temperature rise of the GIS. It optimizes the solution while satisfying the maximum temperature rise constraint of the GIS, achieving synergistic optimization between the structural parameters and thermal performance of environmentally friendly GIS. This method not only reduces busbar material consumption and equipment manufacturing costs but also effectively controls the operating temperature rise of the equipment, improving the operational safety and stability of environmentally friendly GIS, thereby enhancing the economic efficiency and long-term operational reliability of the equipment. Attached Figure Description

[0024] Figure 1 A flowchart of the environmentally friendly GIS multi-dimensional optimization simulation method with multi-physics coupling provided by the present invention; Figure 2 This is a schematic diagram of a three-dimensional model of an environmentally friendly GIS finite element simulation in an embodiment of the present invention; Figure 3 This is a comparison chart of the temperature rise simulation results before and after optimization in an embodiment of the present invention. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of this invention. The embodiments described in this application are merely some embodiments of this invention, and not all embodiments. Based on the spirit of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of this invention.

[0026] Example 1 Embodiment 1 of the present invention proposes an environmentally friendly GIS multivariate optimization simulation method with multiphysics coupling, see reference. Figure 1 ,include: Step 1: Establish a three-dimensional model for environmentally friendly GIS finite element simulation and parameterize the component dimensions and current magnitude; set material parameters and boundary conditions for the three-dimensional model according to the working conditions.

[0027] Specifically, step 1 includes: Step 1.1: Based on the actual or estimated structural dimensions of the environmentally friendly GIS, construct a three-dimensional model for finite element simulation of the environmentally friendly GIS. Step 1.2: To facilitate dimensional parameter adjustment, the established 3D model is appropriately simplified, and the dimensions of five common components affecting temperature distribution—shell outer diameter, shell thickness, shell length, conductor outer diameter, and conductor inner diameter—are set as adjustable parameters. The model simplification is as follows: Figure 2 As shown; Step 1.3: Import the 3D model into the finite element simulation and set the corresponding material parameters and boundary conditions for the 3D model.

[0028] Specifically, step 1.3 includes: Step 1.3.1: Mesh the 3D model. Use hexahedral mesh for the fluid domain and tetrahedral mesh for the solid region. Optimize the mesh quality. For areas with strong geometric nonlinearity, such as sharp corners and component connections, which result in poor mesh quality, refine the mesh by constraining the mesh size to improve mesh quality. Step 1.3.2: Based on the physical constraints required for finite element simulation, add material parameters such as specific heat capacity, heat transfer coefficient, electrical conductivity, and relative permittivity to the 3D model after mesh quality optimization, and set current inlet and outlet surfaces at both ends of the guide rod as excitation sources; Step 1.3.3: Set the temperature correction characteristics for the conductivity of the guide rod and the shell so that the conductivity changes with the temperature, so as to reflect the feedback effect of temperature rise on the conductivity of the conductor under actual conditions.

[0029] Step 2: Based on the 3D model and the set material parameters and boundary conditions, perform multiphysics finite element simulation of electro-thermal-fluid fields; use the Latin hyperisotope sampling method to set multiple combinations of different current magnitudes and size parameters to obtain simulation temperature data of environmentally friendly GIS under multiple working conditions.

[0030] In this embodiment, the Latin hypersampling method is used to select the most representative operating conditions from the simulated temperature data under multiple operating conditions.

[0031] As an optional implementation method, the density of sampling points can be set according to the strength of geometric nonlinear changes in different regions of the 3D model to collect simulated temperature data under multiple operating conditions. For example, in the circuit breaker section of GIS, sampling points should be dense in areas where physical quantities change drastically (such as concentrated electric fields at contact tips or local hot spots); and sparse in areas with gentle changes (such as large flat surfaces of the outer shell or uniform gas areas). Step 3: Using simulated temperature data as a sample data matrix, the sample data matrix is ​​reduced in order to generate a reduced-order data matrix, and the reduced-order data matrix is ​​used to train and generate a reduced-order model.

[0032] Specifically, step 3 includes: Step 3.1: The sample point data matrix is ​​decomposed into a basis vector matrix and an original singular value matrix using the Proper Orthogonal Decomposition (POD) method. Step 3.2: Based on the original singular value matrix, calculate the energy accumulation ratio under different cutoff orders; the specific calculation formula is as follows: ; In the formula, This represents the energy accumulation ratio; The total order; The first singular value in the original singular value matrix Singular values ​​of order; The truncation order; The minimum cutoff order when the energy accumulation ratio is higher than the energy threshold is determined as the final cutoff order; the singular values ​​of the original singular value matrix before the final cutoff order are retained as the low-order singular value matrix; in this embodiment, the energy threshold is 0.95, which means that the modal total energy ratio of the reduced-order model is more than 95% at this time. The energy threshold can be appropriately changed according to the specific working conditions.

[0033] Step 3.3: Extract the first few truncated basis vector matrices from the basis vector matrix as the low-order basis vector matrix; combine the low-order singular value matrix with the low-order basis vector matrix to generate the reduced-order data matrix of the low-order mode. Step 3.4: Use the reduced-order data matrix to train and generate a reduced-order model.

[0034] Step 4: Use the reduced-order model to generate a large number of small-error sample data for SOBO sensitivity analysis, calculate the first-order sensitivity index and total effect index of each input variable, and construct the response surface modeling priority index of each input variable to screen the input variables and obtain the key variables.

[0035] Specifically, step 4 includes: Step 4.1: Based on the actual processing and production capacity of the equipment, the values ​​of the input variables of the reduced-order model are changed multiple times within a predetermined range to generate corresponding small-error sample data. In this embodiment, the number of samples is 448 sets of data. The input variables are the parameterized size parameters, and the value range of the size parameters is as follows: ; Among them, D s H is the inner diameter of the shell; s D is the shell thickness; b1 D is the inner diameter of the conductor; b2 L is the outer diameter of the conductor. s This is the length of the shell.

[0036] Step 4.2: Based on the small error sample data obtained in Step 4.1, perform parameter sensitivity analysis, calculate the variance contribution of each input variable to the output index, and obtain the first-order sensitivity index and the total effect index. The formula for calculating the first-order sensitivity index is as follows: ; In the formula, For input variables; The outputs of the reduced-order model include, for example, the average temperature rise of the entire GIS unit, the maximum temperature rise of the entire unit, and the conductor volume. Indicates in fixed Under the condition that, the conditional expectation obtained by averaging the remaining variables is used to characterize "only by Average response to "decision"; symbol Represents the expectation operation, symbol Variance operation; first-order sensitivity index express The percentage of independent contributions to the overall output variance; The formula for calculating the total effect index is: ; In the formula, the total effect index Indicates and The percentage of all related impacts (including interactions at all levels); Indicates except All variables other than x1 to x5; taking x1 to x5 as an example, E[Y|x ~1 ] represents E[Y|x2,x3,x4,x5].

[0037] Step 4.3: Based on the first-order sensitivity index and total effect index in Step 4.2, consider the magnitude of the individual influence of each variable on the output and the magnitude of the interaction influence with other variables, construct the response surface modeling priority index, and screen out key variables. Specifically, step 4.3 includes: Step 4.3.1: Calculate the linear correlation determination index to characterize the linear explanatory power of each input variable on the output, where the i-th input variable... and the output of the reduced-order model Y The linear correlation index between the input variable and the output of the reduced-order model is calculated by multiplying the covariance between the input variable and the output of the reduced-order model (as the numerator) and the standard deviation of the input variable and the output of the reduced-order model (as the denominator). Specifically, it is expressed as: ; In the formula, For the i-th input variable and output Y The linear correlation index between them; Represents covariance operation; The square of the linear correlation index between the corresponding input variable and the output of the reduced-order model is taken as the corresponding linear correlation determination index, which can then be calculated as follows: ; In the formula, For the i-th input variable and output Y The linear correlation between them determines the index.

[0038] Step 4.3.2: Based on the first-order sensitivity index obtained in Step 4.2 and the linear correlation determination coefficient obtained in Step 4.3.1, define the nonlinear effect index of the i-th input variable; for any input variable, calculate the difference between the first-order sensitivity index and the linear correlation determination index of the corresponding input variable, and take the maximum value between this difference and 0 as the corresponding nonlinear effect index; specifically expressed as: ; In the formula, This indicates the operation of finding the maximum value. N i Let be the nonlinear effect index of the i-th input variable, representing the nonlinear component of the independent influence of the input variable on the output. If N i If it is large, it means that the input variable For output Y The influence has obvious nonlinear characteristics; if N i Conversely, if the size is smaller. Adopt This is to ensure the nonlinear effect index N i It has a non-negative meaning.

[0039] Step 4.3.3: Calculate the difference between the total effect index and the first-order sensitivity index of the corresponding input variable, and take the maximum value between this difference and 0 as the corresponding interaction effect index; specifically expressed as:

[0040] In the formula, Let be the interaction effect index of the i-th input variable; since Indicates input variables For output Y The total impact (including all order interactions) represents Input variables Independent output Y The difference between the two factors can be used to characterize the variable, as they produce an effect. When combined with other input variables, the output Y The resulting interactive effects. If I i If it is large, it indicates that the variable For output Y The influence of a variable stems not only from its independent action but also from its coupling with other variables; if I i The smaller the size, the greater the size.

[0041] Step 4.3.4: Construct the response surface modeling priority index for each input variable. Based on the total effect index, nonlinear effect index, and interaction effect index, define the response surface modeling priority index for the i-th input variable as follows:

[0042] In the formula, e This is the denominator stabilizing factor, which has no specific physical meaning. Its main function is to prevent the denominator from being 0 or too small in the formula. Here, its value is taken as 1 × 10. -8 ; c i For variables The response surface modeling priority index, which characterizes the input variables The necessity and complexity of modeling in response surface methodology (RSM) modeling. This index comprehensively considers the overall influence of a variable on the output response, the nonlinear components of the independent effects of the variable, and the interaction components between the variable and other variables. A larger value indicates a more significant influence of the variable on the current output response, and its influence needs to be expressed through quadratic or interaction terms; therefore, this variable should have a higher modeling priority in RSM modeling. In this index, the outer layer... Used to characterize the overall strength of a variable's influence on the output response. If A smaller value indicates a weaker overall impact of the variable on the output response. Even if its local effect is complex, it should not be given high priority in the response surface model. Therefore, To control variables The importance of fundamentals.

[0043] Step 4.3.5, Priority index for response surface modeling c i Sort the input variables from largest to smallest and filter out those that are most helpful to the output. Y The input variable with the greatest impact on each optimization objective, i.e., the input variable with the highest priority index in response surface modeling, is selected as the key variable. In this embodiment, the key variable selected is the conductor inner diameter. D b1 conductor outer diameter D b2 Length of cylinder L s .

[0044] Step 5: Calculate the busbar material requirements based on the selected key variables; obtain multiple sets of simulation results data by randomly changing the values ​​of key parameters multiple times, and fit the response surface expression; based on the response surface expression, determine the busbar material requirements, average temperature rise of GIS, and maximum temperature rise of GIS under different key variables, construct and solve the optimization objective function based on this, and obtain the optimal key variables.

[0045] Specifically, step 5 includes: Step 5.1: Convert the key variables into vector representations: ; Step 5.2, set the optimization objective as the output, and the output includes △T. max (x), △T avg (x) d b △T max (x) represents the maximum temperature rise of the GIS; △T avg (x); Calculate the busbar material requirements based on key variables; where, busbar material requirements d b The calculation formula is:

[0046] In the formula, Materials used for busbars; The length of the cylinder; D b1 The inner diameter of the busbar; D b2 The outer diameter of the busbar; Step 5.3: Simulations are performed by randomly changing the values ​​of key parameters multiple times to obtain multiple sets of simulation results data for fitting the response surface. The basic expression of the finally fitted response surface is:

[0047] In the formula, This represents the m-th output, where m = 1, 2, 3, specifically the maximum temperature rise ΔT of the GIS. max (x), Average temperature rise of GIS △T avg (x) Busbar materials d b ; This represents the number of key variables, which is set to 3 here. The constant term fitting coefficients are the response surface constants corresponding to the m-th output. The coefficients for the first-order term of the response surface corresponding to the m-th output are: The coefficients for the quadratic term of the response surface corresponding to the m-th output are: The coefficients for the response surface interaction term corresponding to the m-th output are: Let i be the i-th key variable corresponding to the m-th output.

[0048] Based on the fitted response surface, upper and lower limits for the input of key variables are set; the optimization objectives are to minimize the busbar material usage and the average temperature rise of GIS, with the constraint that the maximum temperature rise of GIS does not exceed a predetermined temperature threshold; specifically expressed as follows: ; By weighting two optimization objectives, the multi-objective function is transformed into a single-objective optimization function. Under constraints, a multi-objective genetic algorithm is used to solve the problem, obtaining the optimal values ​​of the key variables within the upper and lower limits of the input, thus completing the optimization design of the flow structure of the environmentally friendly GIS. The optimization results are as follows: Figure 3 As shown.

[0049] After optimization, the maximum temperature rise of the GIS decreased from 133.4℃ to 100.07℃; the volume of busbar materials was reduced by 18.6%, reducing material costs; and the overall average temperature rise decreased from 34.12℃ to 30.80℃.

[0050] Example 2 Embodiment 2 of the present invention provides an environmentally friendly GIS multi-element optimization simulation system with multi-physics coupling, and executes the steps of the environmentally friendly GIS multi-element optimization simulation method with multi-physics coupling described in Embodiment 1, including: The 3D model building module establishes a 3D model for environmentally friendly GIS finite element simulation and parameterizes the dimensions of components and current magnitudes; based on the working conditions, material parameters and boundary conditions are set for the 3D model. The model simulation and data sampling module performs multiphysics finite element simulations of electro-thermal-fluid fields based on the 3D model and the set material parameters and boundary conditions; it sets multiple combinations of different current magnitudes and size parameters to obtain simulation temperature data under multiple operating conditions. The reduced-order model construction module uses simulated temperature data as a sample data matrix, performs reduced-order processing on the sample data matrix to generate a reduced-order data matrix, and uses the reduced-order data matrix to train and generate a reduced-order model. The key variable screening module uses a reduced-order model to generate small-error sample data for sensitivity analysis, calculates the first-order sensitivity index and total effect index of each input variable, and constructs a response surface modeling priority index for each input variable to screen the input variables and obtain key variables. The key variable solving module calculates the busbar material requirements based on the selected key variables; it obtains multiple sets of simulation result data by randomly changing the values ​​of key parameters multiple times, and fits the response surface expression; based on the response surface expression, it determines the busbar material requirements, average temperature rise of GIS, and maximum temperature rise of GIS under different key variables, and constructs and solves the optimization objective function based on this to obtain the optimal key variables.

[0051] This disclosure can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of this disclosure.

[0052] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example—but not limited to—electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination of the foregoing. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0053] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0054] Computer program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing the status information of the computer-readable program instructions to implement various aspects of this disclosure.

[0055] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A multi-physics coupled environmentally friendly GIS multivariate optimization simulation method, characterized in that, include: S1. Establish a three-dimensional model for environmentally friendly GIS finite element simulation, and parameterize the component dimensions and current magnitude; set material parameters and boundary conditions for the three-dimensional model according to the working conditions; S2. Based on the three-dimensional model and the set material parameters and boundary conditions, perform multi-physics finite element simulation of electro-thermal-fluid fields; set multiple combinations of different current magnitudes and size parameters to obtain simulation temperature data under multiple working conditions; S3. Using simulated temperature data as a sample data matrix, the sample data matrix is ​​reduced in order to generate a reduced-order data matrix, and the reduced-order data matrix is ​​used to train and generate a reduced-order model. S4. Use the reduced-order model to generate small-error sample data for sensitivity analysis, calculate the first-order sensitivity index and total effect index of each input variable, and construct the response surface modeling priority index of each input variable to screen the input variables and obtain the key variables. S5. Based on the selected key variables, calculate the busbar material; repeatedly change the values ​​of key parameters to obtain multiple sets of simulation result data, and fit the response surface expression; Based on the response surface methodology, the busbar materials, average temperature rise of GIS, and maximum temperature rise of GIS under different key variables are determined. Based on this, an optimization objective function is constructed and solved to obtain the optimal key variables.

2. The environmentally friendly GIS multi-element optimization simulation method with multi-physics coupling according to claim 1, characterized in that, S3 includes: The sample point data matrix is ​​decomposed into a basis vector matrix and an original singular value matrix; Based on the original singular value matrix, calculate the energy accumulation ratio under different cutoff orders; determine the minimum cutoff order when the energy accumulation ratio is higher than the energy threshold, and use it as the final cutoff order; retain the singular values ​​of the original singular value matrix before the final cutoff order, and use them as the lower-order singular value matrix. Extract the first few basis vector matrices of the final truncation order from the basis vector matrix to form the low-order basis vector matrix; combine the low-order singular value matrix with the low-order basis vector matrix to generate the reduced-order data matrix of the low-order mode. A reduced-order model is generated by training using a reduced-order data matrix.

3. The environmentally friendly GIS multi-element optimization simulation method with multi-physics coupling according to claim 1, characterized in that, S4 include: By changing the values ​​of the input variables of the reduced-order model multiple times within a predetermined range, corresponding small error sample data is generated. Based on small-error sample data, parameter sensitivity analysis is performed to calculate the variance contribution of each input variable to the output index, obtaining the first-order sensitivity index and the total effect index; the formula for the first-order sensitivity index is: ； In the formula, the first-order sensitivity index express The percentage of independent contributions to the overall output variance; Let i be the i-th input variable; This is the output of the reduced-order model; Indicates in fixed Under the given conditions, the conditional expectation obtained by averaging the remaining variables; symbol Represents the expectation operation, symbol Indicates variance operation; Based on the first-order sensitivity index and the total effect index, considering the magnitude of the influence of each variable on the output and the magnitude of the interaction with other variables, a response surface modeling priority index is constructed, and key variables are selected.

4. The environmentally friendly GIS multi-element optimization simulation method with multi-physics coupling according to claim 1 or 3, characterized in that: The process of selecting key variables in S4 is as follows: For any input variable, the covariance between the corresponding input variable and the output of the reduced-order model is used as the numerator, and the product of the standard deviations of the corresponding input variable and the output of the reduced-order model is used as the denominator to calculate the linear correlation index between the corresponding input variable and the output of the reduced-order model. The square of the linear correlation index between the corresponding input variable and the output of the reduced-order model is used as the corresponding linear correlation determination index. Calculate the difference between the first-order sensitivity index and the linear correlation determination index of the corresponding input variable, and take the maximum value between this difference and 0 as the corresponding nonlinear effect index; calculate the difference between the total effect index and the first-order sensitivity index of the corresponding input variable, and take the maximum value between this difference and 0 as the corresponding interaction effect index. Based on the total effect index, nonlinear effect index, and interaction effect index, a priority index for response surface modeling is constructed for each input variable. The input variables are sorted from largest to smallest according to the response surface modeling priority index, and the input variables with the largest response surface modeling priority index are selected as key variables.

5. The environmentally friendly GIS multi-element optimization simulation method with multi-physics coupling according to claim 4, characterized in that: The specific formula for calculating the priority index of response surface modeling is as follows: In the formula, γ i For variables Response surface modeling priority index; Let be the interaction effect index of the i-th input variable; N i Let be the nonlinear effect index of the i-th input variable; For variables The total effect index; Representing variables The first-order sensitivity index; ε This is the stabilizing factor in the denominator.

6. The environmentally friendly GIS multi-element optimization simulation method with multi-physics coupling according to claim 1, characterized in that: The specific representation of busbar material usage in S5 is as follows: ； In the formula, Materials used for busbars; The length of the cylinder; D b1 The inner diameter of the busbar; D b2 This refers to the outer diameter of the busbar.

7. The environmentally friendly GIS multi-element optimization simulation method with multi-physics coupling according to claim 1, characterized in that: The response surface expression in S5 is: In the formula, This indicates the m-th output; This represents the number of key variables, which is set to 3 here. The constant term fitting coefficients are the response surface constants corresponding to the m-th output. The coefficients for the first-order term of the response surface corresponding to the m-th output are: The coefficients for the quadratic term of the response surface corresponding to the m-th output are: The coefficients for the response surface interaction term corresponding to the m-th output are: Let i be the i-th key variable corresponding to the m-th output.

8. The environmentally friendly GIS multi-element optimization simulation method with multi-physics coupling according to claim 1, characterized in that: The process of constructing and solving the objective function in S5 is as follows: Set upper and lower limits for the input of key variables; take minimizing the busbar material consumption and the average temperature rise of GIS as the optimization objectives, and take the maximum temperature rise of GIS not exceeding a predetermined temperature threshold as the constraint condition; transform the multi-objective into a single-objective optimization objective function by weighting the two optimization objectives; solve the optimization objective function under the constraint condition to obtain the optimal key variables within the upper and lower limits of the input of key variables.

9. A multi-physics coupled environmentally friendly GIS multi-element optimization simulation system, using the method described in any one of claims 1-8, characterized in that, include: The 3D model building module establishes a 3D model for environmentally friendly GIS finite element simulation and parameterizes the dimensions of components and current magnitudes; based on the working conditions, material parameters and boundary conditions are set for the 3D model. The model simulation and data sampling module performs finite element simulations of electro-thermal-fluid multiphysics fields based on the 3D model and the set material parameters and boundary conditions. By setting multiple combinations of different current magnitudes and size parameters, simulated temperature data under multiple operating conditions can be obtained. The reduced-order model construction module uses simulated temperature data as a sample data matrix, performs reduced-order processing on the sample data matrix to generate a reduced-order data matrix, and uses the reduced-order data matrix to train and generate a reduced-order model. The key variable screening module uses a reduced-order model to generate small-error sample data for sensitivity analysis, calculates the first-order sensitivity index and total effect index of each input variable, and constructs a response surface modeling priority index for each input variable to screen the input variables and obtain key variables. The key variable solution module calculates the busbar material based on the selected key variables; it obtains multiple sets of simulation result data by randomly changing the values ​​of key parameters multiple times, and fits the response surface expression. Based on the response surface methodology, the busbar materials, average temperature rise of GIS, and maximum temperature rise of GIS under different key variables are determined. Based on this, an optimization objective function is constructed and solved to obtain the optimal key variables.

10. A terminal, comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the steps of the method according to any one of claims 1-8.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the method according to any one of claims 1-8.

Images